Method and system for predicting residual life of wet clutch
Through the improved inverse Gaussian process model, the introduction of random drift and vibration parameters, combined with Bayesian method and Markov chain Monte Carlo algorithm, the limitations of wet clutch degradation analysis and prediction are solved, and accurate time of wet clutch is achieved Life expectation reduces maintenance costs and extends the service life of the equipment.
Patent Information
- Application Number
- CN202411068425.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-08-06
AI Technical Summary
The prior art has limitations in degradation analysis and residual life prediction of wet clutches, which fail to fully consider unit heterogeneity and the importance of online real-time diagnosis, resulting in high experimental costs and prediction inaccuracy.
Using the improved inverse Gaussian process model, random drift and random vibration parameters were introduced, combined with Bayesian method and Markov chain Monte Carlo algorithm, the friction coefficient degradation data was posteriorly sampled, and a model containing random effects was constructed for Real-time prediction of the remaining life of the clutch.
It improves the accuracy and stability of clutch life prediction, reduces experimental costs, realizes real-time health management and predictive maintenance of wet clutch, and extends the reliable service life of the equipment.
Smart Images

Figure CN119066955A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the interdisciplinary field of equipment reliability engineering combined with probability and statistical analysis, and in particular relates to a method and system for predicting the remaining life of a wet clutch. Background Art
[0002] The wet multi-plate clutch is an important power transmission component of the vehicle integrated transmission device and is widely used in various special vehicles. The wet multi-plate clutch mainly relies on the friction pair formed by the friction plate and the release plate to achieve power transmission and interruption, so that high power transmission and control can be easily achieved. However, the cumulative damage caused by the friction and wear effect makes the friction surface gradually thinner and smoother, and loses the expected design function during the service life. Therefore, effective monitoring of the degradation data of the wet clutch, that is, the performance characteristic (PC), can provide a reliable historical information basis for effective predictive maintenance measures and prediction of the remaining useful life (RUL). Through accurate prediction methods, the service life and reliability of the wet clutch can be extended, and the working cost during the use stage can be reduced.
[0003] Traditional reliability analysis based on life data incurs higher costs in terms of acquisition and implementation of available data. Compared with life data, PC reflects the evolution of clutch performance more accurately and also contains more life information. Degradation analysis, including degradation modeling and parameter estimation, has proven to be an important toolkit, especially for equipment with limited testing time and sample size. In addition, we can use failure mechanisms to guide the modeling of product degradation processes, conduct reliability analysis and predict RUL, which is instructive for condition-based maintenance. A suitable degradation model is essential for clutch degradation characteristics and reliability representation. Physics-based methods usually require clear mathematical models to quantitatively characterize the degradation behavior of the system, which is almost impossible for complex clutch systems. Stochastic process-based degradation modeling has become a more promising option due to its ability to capture random dynamics and unit heterogeneity (usually including random drift and random vibration) in the degradation process. In particular, the Inverse Gaussian process (IG) has been particularly prominent in recent studies and has been proven to be an attractive degradation process model.
[0004] Although many scholars have achieved certain research results in exploring the degradation of clutch working characteristics and remaining life prediction using finite element simulation and physical model methods. However, there is still a large research gap in the degradation analysis and RUL prediction of complex wet clutch systems based on random processes: many studies do not fully consider the unit heterogeneity between the same and different clutch devices, including long-term trends and short-term effects; in addition, most existing studies use a large amount of offline data to initialize the model to obtain higher model performance, ignoring the importance of online real-time diagnosis of equipment, which is unaffordable for expensive clutch systems. Therefore, there is an urgent need for an effective life prediction model based on degradation data to guide the degradation prediction and real-time prediction of RUL of the clutch system, guide the implementation of condition-based maintenance strategies and health management, and ensure the efficient and reliable operation of the clutch during the service phase. Summary of the invention
[0005] In order to solve the above technical problems, the present invention proposes a method and system for predicting the remaining life of a wet clutch, which can solve the limitations of the prior art in terms of clutch degradation analysis and remaining life prediction.
[0006] To achieve the above object, the present invention provides a method for predicting the remaining life of a wet clutch, comprising:
[0007] Acquiring clutch data to be predicted;
[0008] Inputting the clutch data to be predicted into an improved IG process model to obtain a clutch remaining life prediction value, wherein the improved IG process model is trained by a training set, and the training set is friction coefficient degradation data;
[0009] The improved IG process model is: introducing random drift parameters and random vibration parameters to modify the parameter structure of the IG process model.
[0010] Optionally, obtaining the friction coefficient degradation data includes:
[0011] Using sensors, obtaining initial friction coefficient degradation data of the clutch;
[0012] The initial friction coefficient degradation data is subjected to an outlier elimination process to obtain the friction coefficient degradation data.
[0013] Optionally, the improved IG process model further includes:
[0014] The prior distribution and posterior distribution of the parameters are defined by using the Bayesian method; the IG process model after the structural parameters are modified is sampled by using the MCMC algorithm to obtain the parameter statistics to be estimated of the improved IG process model.
[0015] Optionally, the random drift parameter is μ, and its probability density function is:
[0016]
[0017] The random vibration parameter is λ, and its probability density function is:
[0018]
[0019] Among them, ω, κ, η and γ are the hyperparameters to be estimated in the model, B is the normalization factor, is the gamma function, μ is the random drift parameter, λ is the random vibration parameter, and φ is the probability density symbol of the standard normal distribution.
[0020] Optionally, the IG process model after the structural parameters are modified is:
[0021]
[0022] Where Λ(t) is the time-dependent shape function; y t =Y(t), which is the degraded observation value at time t.
[0023] Optionally, before inputting the clutch data to be predicted into the improved IG process model, the method further includes:
[0024] Obtain the latest degradation observation data;
[0025] Based on the latest degradation observation data, degradation data of a preset time window length is set.
[0026] Optionally, obtaining the predicted value of the remaining life of the clutch includes:
[0027] predicting a degradation state based on the improved IG process model;
[0028] It is determined whether the degradation state exceeds a preset fault threshold, and based on the determination result, a predicted value of the remaining life of the clutch is obtained.
[0029] The present invention also provides a wet clutch remaining life prediction system, comprising: a bench test module, a data processing module, a numerical modeling module, and a performance prediction module;
[0030] The bench test module is used to design a cyclic engagement test of the clutch system under actual vehicle conditions, and uses sensors to collect all COF degradation data of the clutch system from the time it is put into use to the end of the test;
[0031] The data processing module is used to perform outlier elimination processing on the COF degradation data;
[0032] The numerical modeling module is used to construct an improved IG process model of a modified parameter structure including random drift and random vibration;
[0033] The performance prediction module is used to obtain a predicted value of the remaining life of the clutch based on an improved IG process model.
[0034] Compared with the prior art, the present invention has the following advantages and technical effects:
[0035] The model in this invention aims to improve the accuracy and stability of clutch RUL prediction by using small sample degradation data, while reducing the reliance on high experimental costs. By effectively analyzing the historical degradation data of the clutch and combining random effects, the improved IG process model shows excellent prediction accuracy. This invention helps to extend the reliable service life of the clutch, reduce maintenance costs, and provide effective support for vehicle health management and predictive maintenance. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0037] Figure 1 is a flow chart of a method for predicting the remaining life of a wet clutch according to an embodiment of the present invention;
[0038] Figure 2 This is a comparison chart between the actual value and the predicted value of the remaining life of the clutch according to an embodiment of the present invention. DETAILED DESCRIPTION
[0039] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0040] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0041] The present invention aims to solve the limitations of the prior art in clutch degradation analysis and remaining life prediction. First, the present invention accurately extracts the key PC parameter friction coefficient (COF) that can characterize clutch performance degradation from the degradation data. Secondly, considering the nonlinear characteristics and complex dynamic behavior of the clutch system, an improved IG process model based on small sample degradation data is developed. By making full use of historical data mining and learning the degradation characteristics of different time stages, the accuracy and stability of the prediction are improved to achieve real-time prediction of the clutch RUL.
[0042] The present invention proposes a method for predicting the remaining life of a wet clutch. Figure 1 As shown, specifically including:
[0043] Acquiring clutch data to be predicted;
[0044] The clutch data to be predicted is input into the improved IG process model to obtain the predicted value of the remaining life of the clutch. The improved IG process model is trained by a training set, which is the friction coefficient degradation data.
[0045] Furthermore, obtaining friction coefficient degradation data includes:
[0046] Using sensors, obtaining initial friction coefficient degradation data of the clutch;
[0047] The initial friction coefficient degradation data is processed by eliminating outliers to obtain the friction coefficient degradation data.
[0048] Specifically, a cyclic engagement test of the clutch system under actual vehicle conditions is designed, and sensors are used to collect all friction coefficient degradation data of the clutch system from the time it is put into use to the end of the test. In order to control the influence of irrelevant variables and ensure the validity of the degradation data, the clutch must be forced to cool after each engagement test.
[0049] The basis for judging the end of the clutch life cycle test is: the first clutch engagement time is t1, and the nth engagement time is tn; if tn ≥ 5*t1, the full life cycle test is over. If smoke (oil vaporization) or abnormal noise occurs during the experiment, the full life cycle test can also be ended.
[0050] The test interval of the clutch cycle test is 100 cycles, that is, the COF degradation value is collected every 100 engagement cycles.
[0051] In view of the inherent nonlinearity and complexity of the clutch system, outliers in the COF degradation data are preliminarily eliminated through data preprocessing and trend smoothing is performed as the input PC parameters of the subsequent model.
[0052] Furthermore, the improved IG process model is:
[0053] The parameter structure of the IG process model is modified by introducing random drift parameters and random vibration parameters;
[0054] The Bayesian method is used to test the IG process model after the structural parameters are modified to obtain the improved IG process model.
[0055] Specifically, considering the unit heterogeneity of the clutch system, an improved IG process model IG_RDV with a modified parameter structure including two random effects (i.e., random drift and random volatility) is constructed. Then, the model-related sub-modules are configured and the prior parameters are set, including the introduction of random effects and their hyperparameters, the selection of shape functions, etc.; then, the sample sampling of the posterior predictive test of the model is realized through the Bayesian method and the Markov Chain Monte Carlo algorithm (MCMC), and the parameter estimation results are obtained, and the fitting ability of the model using degraded data is preliminarily verified.
[0056] Random drift refers to the uncertainty of the overall trend of a random process over time, which is a long-term characteristic; random vibration represents the uncertainty of the instantaneous fluctuation of the random process, which is a short-term irregular change superimposed on the long-term trend. Introducing random drift and random vibration into the IG process model at the same time can better characterize the uncertainty of the performance degradation of the complex clutch system in the long-term and short-term process. When using Bayesian MCMC for model parameter inference, the first 800 samples of the simulation process are discarded through the 'burned-in' operation (so that the sampling results are stable in the target distribution area), and the last 1500 samples are used as representatives of the degradation samples.
[0057] The improved IG process model with random effects mainly includes the following steps: For the simple IG process IG(μΛ(t),λΛ(t) 2 ),μ,λ>0 (abbreviated as IG_Sim), when no random effect is introduced, μ and λ are constants, and its probability density function (PDF) is:
[0058]
[0059] Where y = Y(t), Λ(t) is the time-dependent shape function, usually in the form of a linear function, a power law function, etc. According to the degradation characteristics of COF data, the shape function Λ(t) = t q .
[0060] When random effects are introduced, the simple IG process parameter structure is modified to IG(μΛ(t),μ 3 Λ(t) 2 / λ), and let μ~TN(ω,κ -2 ),λ~Gamma(η,γ) introduce random drift and random vibration respectively, where TN and Gamma are truncated normal distribution and gamma distribution respectively, and ω,κ,η and γ are hyperparameters to be estimated in the model respectively. The PDFs of random effect parameters μ and λ are:
[0061]
[0062] Where B is the normalization factor, B = 1-Φ(-κω); is the gamma function.
[0063] The PDF of the IG model IG_RDV after modifying the parameter structure is:
[0064]
[0065] Then the joint posterior distribution of the model can be expressed as:
[0066] π(ω,κ,η,γ,Λ(t)|Y)∝π(ω,κ,η,γ,Λ(t))L(Yω,κ,η,γ,Λ(t))
[0067] Among them, π(·) and L(·) are the prior distribution and likelihood function of the parameters to be estimated, respectively.
[0068] Furthermore, before inputting the clutch data to be predicted into the improved IG process model, the following steps are further included:
[0069] Obtain the latest degradation observation data;
[0070] Based on the latest degradation observation data, set the degradation data of the preset time window length.
[0071] Furthermore, obtaining the predicted value of the remaining life of the clutch includes:
[0072] Predict degradation status based on improved IG process model;
[0073] Determine whether the degradation state exceeds a preset fault threshold, and obtain a predicted value of the remaining life of the clutch based on the determination result.
[0074] Specifically, based on the established IG process model and the concept of the first passage time of equipment failure, the RUL iterative prediction of the clutch system is realized by using the time window of degradation data (i.e., the preprocessed COF data). That is, through the iterative degradation data time window, the model infers the degradation of the clutch system and estimates the parameters according to the latest degradation information, and finally obtains the predicted RUL based on the posterior sample distribution.
[0075] The steps for RUL prediction using the degradation time window are as follows:
[0076] Let Y i (t 1:k ) = {y i (t1), y i (t2), …, y i (t k )} represent the observed values of the degradation data of the clutch system. According to the preset failure threshold D and the definition of the first passage time, the remaining life of the system at time t is R t = inf{d > 0: Y(t + d) ≥ D|Y(t) < D}. It is difficult to directly obtain the analytical solution of the target distribution, so the Bayesian MCMC method is used to obtain an approximation of the objective function. The specific steps for RUL calculation are as follows:
[0077] (1) Obtain the latest degradation observation data (COF) at the current time t k , and select the degradation data Y i (t k-l:k ) = {y i (t k-l ), y i (t k-l+1 ), …, y i (t k )} with a preset time window length of l;
[0078] (2) Input the observed data in the time window into the established degradation process model, and obtain the posterior predictive distribution and estimated parameters based on the current information;
[0079] (3) Use the newly estimated parameters and the degradation model to predict the future degradation state, that is, y i (t k + d) = y i (t k + d - 1) + Δy i (t k + d - 1), d ≥ 1;
[0080] (4) Determine y i (t k+d) whether it exceeds the preset fault threshold D. If not, continue to perform degradation prediction based on the current information; if so, the fault time and RUL are t k +d and d.
[0081] Among them, y i (t k +d) represents the predicted future time t k +d degenerate value; Δy i (t k +d-1) represents the predicted future time t k +d-1 degradation increment; d is the degradation model using the current time t k The RUL prediction results obtained from the degradation data time window.
[0082] The fitting performance and complexity of the model were evaluated based on the maximum log-likelihood value and Akaike information criterion (AIC); the mean absolute error (MAE) was used to evaluate the predictive performance of the model.
[0083] The calculation formula of AIC is: Where k is the number of parameters in the model, is the maximum log-likelihood value of the model. The smaller the AIC, the better the model's goodness of fit and complexity. is the value of the likelihood function L at the maximum likelihood estimate, which indicates the maximum probability or probability density of the observed data (degraded data) under the optimal parameter estimate. In order to illustrate the fitting excellence of the proposed model, Table 1 shows the maximum log-likelihood value and AIC value obtained under different IG process models using the COF data obtained from the clutch degradation test.
[0084] Table 1
[0085]
[0086] The present invention uses the mean absolute error (MAE) indicator to evaluate the RUL prediction performance of the model, and its expression is as follows:
[0087]
[0088] Where N represents the number of prediction time points from the prediction start time to the actual failure time; It represents the average value of the sample RUL during prediction; D act is the actual value of RUL.
[0089] The present invention also provides a wet clutch remaining life prediction system, comprising: a bench test module, a data processing module, a numerical modeling module, a performance prediction module and a data evaluation module;
[0090] The bench test module is used to design the cyclic engagement test of the clutch system under actual vehicle conditions, and uses sensors to collect all COF degradation data of the clutch system from the time it is put into use to the end of the test;
[0091] The data processing module preliminarily removes outliers in the COF degradation data and performs trend smoothing through data preprocessing;
[0092] Numerical modeling module for building improved IG process models for modified parameter structures including random drift and random vibration;
[0093] The performance prediction module uses the degradation data time window to implement the clutch system RUL iterative prediction and obtain the clutch remaining life prediction value;
[0094] Data evaluation module: The fitting performance and complexity of the model are evaluated based on the log-likelihood value and Akaike information criterion, and the mean absolute error (MAE) is used to evaluate the prediction performance of the model.
[0095] In order to fully demonstrate the superiority of the RUL prediction model of the present invention, the MAE index of the RUL prediction results with and without considering the random effect is calculated on a clutch system COF degradation process data set, as shown in Table 2:
[0096] Table 2
[0097]
[0098]
[0099] In order to intuitively show the error trend between the RUL predicted value and the true value, the following is given: Figure 2 The results are compared as shown. It can be seen that the present invention can accurately predict the RUL of the clutch system in real time, so as to timely grasp the working quality and health status of the clutch, and provide a scientific basis for the health management and predictive maintenance of the clutch during service. This technical solution helps to extend the service life of the clutch, reduce maintenance costs, and ensure the safe and stable operation of the vehicle.
[0100] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A method for predicting the remaining life of a wet clutch, characterized in that: include: Acquiring clutch data to be predicted; Inputting the clutch data to be predicted into an improved IG process model to obtain a clutch remaining life prediction value, wherein the improved IG process model is trained by a training set, and the training set is friction coefficient degradation data; The improved IG process model is: introducing random drift parameters and random vibration parameters to modify the parameter structure of the IG process model.
2. A method for predicting the remaining life of a wet clutch according to claim 1, characterized in that: Acquiring the friction coefficient degradation data includes: Using sensors, obtaining initial friction coefficient degradation data of the clutch; The initial friction coefficient degradation data is subjected to an outlier elimination process to obtain the friction coefficient degradation data.
3. A method for predicting the remaining life of a wet clutch according to claim 1, characterized in that: The improved IG process model also includes: The prior distribution and posterior distribution of the parameters are defined by using the Bayesian method; the IG process model after the structural parameters are modified is sampled by using the MCMC algorithm to obtain the parameter statistics to be estimated of the improved IG process model.
4. A method for predicting the remaining life of a wet clutch according to claim 3, characterized in that: The probability density function of the random drift parameter μ is: The probability density function of the random vibration parameter λ is: Among them, ω, κ, η and γ are the hyperparameters to be estimated in the model, B is the normalization factor, is the gamma function, μ is the random drift parameter, λ is the random vibration parameter, and φ is the probability density symbol of the standard normal distribution.
5. A method for predicting the remaining life of a wet clutch according to claim 4, characterized in that: The probability density function of the IG process model after the structural parameters are modified is: Where Λ(t) is the time-dependent shape function; y t =Y(t), which is the degraded observation value at time t.
6. A method for predicting the remaining life of a wet clutch according to claim 1, characterized in that: Before inputting the clutch data to be predicted into the improved IG process model, the method further includes: Obtain the latest degradation observation data; Based on the latest degradation observation data, degradation data of a preset time window length is set.
7. A method for predicting the remaining service life of a wet clutch according to claim 1, characterized in that: Obtaining the predicted value of the remaining life of the clutch includes: predicting a degradation state based on the improved IG process model; It is determined whether the degradation state exceeds a preset fault threshold, and based on the determination result, a predicted value of the remaining life of the clutch is obtained.
8. A wet clutch remaining life prediction system, characterized in that: include: Bench test module, data processing module, numerical modeling module, performance prediction module; The bench test module is used to design a cyclic engagement test of the clutch system under actual vehicle conditions, and uses sensors to collect all friction coefficient degradation data of the clutch system from the time it is put into use to the end of the test; The data processing module is used to remove abnormal values from the friction coefficient degradation data; The numerical modeling module is used to construct an improved IG process model of a modified parameter structure including random drift and random vibration; The performance prediction module is used to obtain a predicted value of the remaining life of the clutch based on an improved IG process model.
Citation Information
Patent Citations
Method and device for predicting service life of vehicle transmission part and vehicle
CN116026607A
Method and system for predicting remaining useful life of industrial equipment, and electronic device
WO2021000556A1
Cited By
Large film evaporator fault prediction method and system
CN120596897A
A large thin film evaporator fault prediction method and system
CN120596897B
